Just get a desk job

So the idea is to make a way for the arms to get up onto tables in order to be more useful.

Basically,

1. the robot carries around an extra clamp device. Maybe on its shoulder, maybe somewhere else.
2. It then mounts the clamp (somehow) onto a surface.
3. Then it grabs the surface and disconnects itself from the base.
4. It then places itself up on the clamp.
5. From here it has a good position to access the local desk area.

Then the other arm climbs up there (not shown) so that we have two arms on a desk. And when the job is done, it climbs down and goes to the next location.

And then comes 12 months of social influencers predicting the end of the world for everyone with a desk job.

But, before that… let’s start with the pressing issues.

Issues

Weight and grip

Robot arms are generally quite heavy. There might even be no prior cases of a working solution or even an attempt at this (too bold a statement).

There are now smallish actuators that can lift pretty heavy things. But it's still an assumption that they would be practical in this use case.

The gripper grabbing the table is also a bit of a red flag. A lot can go wrong regarding different surfaces. It would make more sense that the gripper connects with the clamp itself. Then the clamp (or various different clamps) takes full responsibility for gripping various surfaces. Rather than the gripper potentially sacrificing its dexterity for this function.

This would also allow the arm to test the clamp for a stable connection before using it.

Cables and hardware drivers

It's not really decided how the power will go.

It's likely that batteries and extra power connections will stay in the base. The motor-driving electronics might also be kept there.

That means some cable management from the base up to the arm on the desk. It’s going to be messy.

Software, sensors, control, etc.

Yes, well. One does not simply assume this is easy.

But, I was tempted to put a little black box on the robot with “AI” to represent “Assumably Ignorant” of the difficult software solution needed to make it happen.

That brings me to the…

Business model

which is also coming soon.

Progress Sketches

Here’s some working material.

Here I traced the initial CAD drawing back into the Remarkable tablet in order to use it as a reference.

Isometric background templates really make a big difference for sketching.

Locking interface

The arm and clamp would have some kind of … locking interface.

Perhaps this could be activated by the help of the other arm.

here is some sketches that kinda show the behaviour.

This picture is particularly terrible.

concept CAD

What size clamp dimensions?

How will the locking interface work?

The way to make a good clamp and locking interface is to contract someone with experience.

This will be scoped out in another study. Untitled

Before that: it needs to be sorted out if the concept is practical at all.

This basic concept serves purely as a communication tool.

It's just a simple clamp with a locking mechanism on several sides to make it easy to play around with.

In reality, the final work is likely to be a number of general and specialised attachments. Not just a one-size that tries to do every job.

Re-dimensioning

I update the model a little for this post.

At this point, there is not a huge value in getting the size exactly right.

But, still, I did some basic resizing of the base. (see last post on size) However, the “size box” is now on its side. Regarding “easy storage”, it's not clear if this is important at all.

Upcoming work on the business model suggests the robot may never really be stored at all.

I also managed to fix the axis problem on the big arm as mentioned in the previous post.

The arm now sits higher on the base.

Where and how the arm is to be mounted will need to be workshopped later.

This will do for now.

The files:

Are here on grabcad